Comparison
Pigsty vs managed RDS
| Aspect | Cloud-managed RDS | Pigsty |
|---|---|---|
| Cost | Pay-per-hour; expensive long-term | Pay hardware cost once |
| PostgreSQL version | Usually 6–18 months behind upstream | Tracks upstream; 17 by default |
| Extensions | Dozens | 340+ |
| Data sovereignty | Subject to cloud provider | Fully in your hands |
| Customization | Limited | Unlimited |
| Lock-in risk | High | None |
Pigsty vs Kubernetes operators (CloudNativePG, etc.)
| Aspect | K8s operator | Pigsty |
|---|---|---|
| Infrastructure assumption | Existing K8s cluster | Bare metal / VMs |
| Operational mindset | Cloud-native / container scheduling | Traditional "pet" nodes |
| Monitoring stack | Bring your own | Delivered by default |
| Extension ecosystem | Build your own image | Precompiled, one-stop |
| Best fit | Short-lived, elastic databases | Long-running production DB |
Pigsty vs plain PostgreSQL
Plain PostgreSQL is just a database process. In production you also need:
- A high-availability solution (Patroni + etcd)
- A connection pooler (pgBouncer)
- A load balancer (HAProxy)
- A backup tool (pgBackRest)
- Monitoring (Prometheus / Grafana / exporter)
- Logging, alerting, config management, …
Pigsty = PostgreSQL + all of the above + opinionated wiring, instead of making you glue the pieces yourself.